Page 63 - BH-3-3

Brain & Heart AI in biomarker discovery for CVDs

Table 1. Example of various classes of CVD biomarkers a The intricacies of human diseases and the inherent
shortcomings of diagnostics reliant on single markers
Biomarker Examples Application highlight the need for a more comprehensive approach.
type
Proteins BNP, NT-proBNP, Heart failure, myocardial A multi-analyte approach can provide significant
troponin infarction advantages over traditional single-analyte strategies
12
Non-coding miRNAs (e.g., Early markers for acute for disease diagnosis and prognosis. Integrating these
RNAs miR-208, miR-499) coronary syndromes biomarkers into clinical practice presents significant
Cellular DNA Circulating cell-free Reflects tissue injury challenges, necessitating standardized protocols and
DNA comprehensive validation studies. An integrated approach
Metabolic Lipoprotein (a), Metabolic changes, promises to transform precision medicine by offering
markers homocysteine atherosclerosis deeper insights into disease mechanisms, enabling
Inflammatory CRP, IL-6 Inflammation associated early detection, and facilitating personalized treatment
markers with CVDs strategies. The combined use of circulating miRNAs with
a
Note: Data were obtained from ref. 6 other biomarkers offers a valuable path for thorough disease
Abbreviations: BNP: B-type natriuretic peptide; CRP: C-reactive management. Furthermore, the emergence of AI and
protein; CVDs: Cardiovascular diseases; NT-proBNP: N-terminal pro machine learning (ML) could markedly boost these efforts.
B-type natriuretic peptide; IL: Interleukin; IL-6: Interleukin-6.
Integrating AI into medical research could revolutionize
the diagnosis, optimization of treatment strategies, and
miRNAs and other novel biomarkers reflects the dynamic
nature of research in this field, creating new pathways for refinement of prognosis predictions. AI’s ability to process
early diagnosis and tailored therapeutic approaches. Omics complex, multidimensional data significantly enhances
technologies have dramatically transformed biomarker the accuracy of early detection and the personalization of
12
discovery. These advanced platforms facilitate high- treatment plans.
10
throughput profiling of a broad spectrum of biological AI is profoundly transforming various fields, including
molecules across diverse cellular and tissue states. Capable healthcare, where its impact on biomarker discovery for
of measuring millions of features, including genotypes, CVDs is notable. Biomarkers are essential for diagnosing,
epigenetic states, and the levels of RNAs, proteins, and predicting, and monitoring diseases, and AI’s capacity
metabolites, omics technologies have broadened the to process large datasets and identify complex patterns
horizon for identifying novel biomarkers. significantly enhances these processes.
However, despite these advanced capabilities, the AI technologies, such as natural language processing,
pathway from potential biomarkers to clinically validated ML, rule-based expert systems, robotic process
ones present significant challenges. Only a few biomarkers automation, and physical robots, offer distinctive
11
have achieved the level of robustness required for definitive capabilities ranging from predictive modeling and disease
analytical and clinical application, emphasizing the need detection to improving surgical accuracy and automating
for rigorous validation processes to ensure their efficacy administrative functions. Integrating AI into healthcare is
and reliability in clinical settings. Transitioning biomarkers poised to enhance diagnostic precision, support decision-
from research to clinical use requires ongoing innovation making, and refine treatment plans, which could reduce
and rigorous validation. This process is critical to ensuring medical errors and elevate patient outcomes. 13
that new discoveries translate into tangible benefits for The complexity of human diseases, particularly CVDs,
patient care, thereby maximizing the impact of omics necessitates a sophisticated approach to biomarker
technologies in cardiovascular health. discovery. Traditional single-marker diagnostics are often
3. Revolutionizing biomarker discovery for insufficient due to the multifaceted nature of CVDs, which
arise from genetic, environmental, and lifestyle factors. AI’s
CVDs with artificial intelligence (AI) ability to sift through extensive datasets – including genetic
As previously mentioned, despite the potential of information, electronic health records, and lifestyle data
biomarkers, particularly miRNAs, as diagnostic, prognostic, – allows for the discernment of patterns and correlations
and therapeutic tools in CVDs, several challenges hinder indicative of potential biomarkers. This capability is crucial
their clinical application. A major obstacle in miRNA for developing predictive models that improve disease
biomarker discovery and validation is the biological progression analysis and enable personalized treatments.
complexity of disease pathogenesis. Technical issues, Integrating omics with AI technologies furthers the
11
such as the lack of standardization in sampling, processing, identification of risk markers for diseases such as heart
and storage of samples, further complicate this process. failure, assists in monitoring care, determines prognosis, and

Volume 3 Issue 3 (2025) 3 doi: 10.36922/bh.8442